This study was part of a randomized clinical trial comparing the central nervous system effects of the two vital organ-support methods used in infant cardiac surgery: total circulatory arrest and low-flow cardiopulmonary bypass. The extent to which visual-spatial deficits are (1). associated with surgical and perioperative variables, (2). attributable to visual-perceptual, motor control, or metacognitive deficits, and (3). associated with adaptive difficulties at home or school was evaluated. The subjects were 155 8-year-old children with D-transposition of the great arteries who underwent the arterial switch operation before 3 months of age. As part of a comprehensive evaluation, the Rey-Osterrieth Complex Figure (ROCF) was administered. ROCF copy productions were classified as having a Basal Organization Level of 1 (low) or 2 or greater. A five-category clinical rating was also assigned. More than half of the children in the cohort (52%) had copy productions scored at Level 1, more than twice the expected frequency. The risk of having a low score was not associated with vital organ support method or other surgical variables. On the basis of comparisons of the relative fits of nested logistic regression models, poor visual-perceptual abilities were more predictive of having a Level 1 score than either motor control or metacognitive deficits. Children with poor copy production scores had lower mathematics scores, but not lower reading scores or poorer parent and teacher ratings of adaptive competence. The percentage of children receiving remedial school services was associated with ROCF clinical rating, ranging from 58% in the worst category to 8% in the best category. Visual-spatial deficits are common among children after infant heart surgery and seem to reflect visual-perceptual rather than motor control or metacognitive deficits. In addition, these deficits do not seem to be clearly associated with the intraoperative methods or postoperative events evaluated.